Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
/plugin marketplace add LeeJuOh/claude-code-zeronpx agentmods add plugins/leejuoh/claude-code-zero/rubber-duck-tutorgit clone --depth 1 https://github.com/LeeJuOh/claude-code-zeroGrade A, and why
rubber-duck-tutor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
{
"name": "rubber-duck-tutor",
"description": "Rubber duck tutor that keeps you sharp across the whole AI-coding lifecycle, not just after code lands: /duck-prebuild before you ask AI to build, /duck-verify and /duck-review once it's written and before you ship, /duck-orient when you're new to a codebase, plus /duck for auto-detect. /coach is its teaching sibling (manual-invocation only): /coach <topic> explains and drills a concept, /coach <file> or a bare /coach right after generating code dissects it section by section, and /coach alone picks up an unresolved gap from a past duck session and teaches it — resolved only once you pass the exercise, not by saying you understand. Ship-point hooks (git push, gh pr create, glab mr create) confront you inline — targeting the highest-risk change you didn't actually discuss, retrying an unresolved gap from a past session, or demoting to a non-blocking scoreboard after three ignored confrontations in a row — and every fire/answer/ignore is logged so you can see whether it's working. Doesn't referee code quality or plan design — that's /code-review and /grilling's job; duck only checks whether you understand, and never teaches — that's coach's job.",
"author": { "name": "LeeJuOh" },
"license": "MIT",
"keywords": ["rubber-duck", "tutor", "learning", "code-review", "rubber-stamping", "skill-development"]
}
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 8 lines scan A 60a6f3fab738
rubber-duck-tutor is a plugin published in the GitHub repository LeeJuOh/claude-code-zero (51 stars, last pushed 2d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other plugins, from other repositories
explore-claude-code
Learn Claude Code by exploring it as it was designed - interactive IDE-style docs for commands, MCP, skills, CLAUDE.md and more.
my-plugin
Learn Claude Code by exploring it as it was designed - interactive IDE-style docs for commands, MCP, skills, CLAUDE.md and more.
skills
Skill discovery, installation, and authoring — find skills with scout, create new universal skills with skill-builder.
book-to-skill
Turns a Markdown knowledge source (book, manual, course, transcript) into a Claude skill: extracts the transferable method with source anchors, validates every unit against the source, assembles a thin SKILL.md plus reference sheets, and measures the result with evals.
kiln
Multi-model software creation pipeline. A conversation becomes running, tested software — an interactive brainstorm, then an autonomous build: GPT-5.6 writes the code, Claude shapes the plans and interfaces, and each family inspects the work of the other.
coaltipple
CoalTipple — model/effort router for Claude Code. Delegates a task you CAN do but is large+cheap DOWN to a lower tier to save tokens; escalates a task beyond the current tier UP for quality — up to Fable 5, the top rung (a consent-gated, real-money spawn). Two-knob routing (tier × effort, effort-before-tier), an alias.